You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As discussed in google/Xee#153, I'll put together a PR with the core functionality of loading Earth Engine data into a Dask dataframe via a top-level read_ee function.
We settled on dask-geopandas as an optional dependency to enable geospatial processing on deferred dataframes, but I'm leaning towards omitting that from the MVP until geopandas/dask-geopandas#285 is merged. The alternative would be to pin to an older Dask before the incompatible query planning system was added or disable query planning in the Dask configuration. Given that there's a fix in progress, I'm in favor of just postponing that feature until it's released and then pinning to that.
That doesn't stop us from optionally using GeoDataFrames for the underlying chunks in the Dask dataframe, and if a user needs to run geospatial operations without computing they can always set their own Dask config and wrap the read_ee output in from_dask_dataframe.
Let me know if you have strong feelings on that @alxmrs - otherwise I'll move ahead with that plan.
The text was updated successfully, but these errors were encountered:
As discussed in google/Xee#153, I'll put together a PR with the core functionality of loading Earth Engine data into a Dask dataframe via a top-level
read_ee
function.We settled on
dask-geopandas
as an optional dependency to enable geospatial processing on deferred dataframes, but I'm leaning towards omitting that from the MVP until geopandas/dask-geopandas#285 is merged. The alternative would be to pin to an older Dask before the incompatible query planning system was added or disable query planning in the Dask configuration. Given that there's a fix in progress, I'm in favor of just postponing that feature until it's released and then pinning to that.That doesn't stop us from optionally using GeoDataFrames for the underlying chunks in the Dask dataframe, and if a user needs to run geospatial operations without computing they can always set their own Dask config and wrap the
read_ee
output infrom_dask_dataframe
.Let me know if you have strong feelings on that @alxmrs - otherwise I'll move ahead with that plan.
The text was updated successfully, but these errors were encountered: